When Thomas Gray inked the phrase, "Ignorance is Bliss, 'tis folly to be wise," I don’t think he considered how best to manage data in our present-day data Armageddon. If you are a data manager and you adhere to the "ignorance is bliss" school of thought, I would recommend that you refresh your resume immediately!
I have spoken with too many people who have no idea of what is to come concerning the world’s rapid and exponentially growing data. Believe it or not, I talked to a person at the Supercomputing show in Seattle who said they are actually moving all their data to disk and neglecting the tremendous, inherent values and benefits (low cost, high capacity and performance, to name a few) of tape. As their data doubles each year, which he said it does, the plan is to continue adding more disk... Really? In his case, I believe he really thinks ignorance is bliss. I offered to share with him how customers with hundreds of terabytes to hundreds of petabytes are managing data with intelligent file systems and using both tape and disk in cost efficient ways and he refused to listen because his ignorance has caused him to believe that "tape is dead". Granted, I don’t hear this very often anymore because the HPC community, as a whole, is paving the way for a cost-effective tape-based storage concept we will discuss later, called "Active Archive".
First, I want to address the ignorance of the individuals who have sipped the "tape is dead" Kool- Aid from certain disk vendors over the past 10 years. Growing up as a teenager in the great state of Texas, I listened to AM radio in my first pickup truck. (Yes, all it had was an AM radio!) Anyway, one of my favorite radio talk shows was Mr. Earl Pitts, who addressed controversial topics and would start by sharing his straightforward opinion on them by saying (insert Texas accent)"Ya know what makes me sick, you know what makes me so angry I could spit?"… or something along those lines. (http://www.youtube.com/watch?v=4DDhrRooNp4) Then he would talk about something that is usually contradictory to the American way since he was a patriot who was always watching out for our true, red-blooded American values. Well, I feel sort of like Earl when someone tells me that they think that tape is of no value, which simply shows their ignorance. I want to say “you know makes me sick, you know what makes me so angry I could spit?".....Ignorance! He would always end his lesson on values and truth by saying “Wake up America!” Well, when someone tells me “tape is dead”, I want to grab them, shake them and say “Wake up!”
The reality today, regarding data storage, is that it is not folly to be wise and it is not bliss to be ignorant. Wake up Storage Admins! I have to admit that the number of people I talk to around the country at trade shows, in meetings, etc., are awake and aware of the ever present danger of data explosion. So, needless to say, my blood pressure stays in check and I don’t get angry as often. I try to keep things in perspective and just assume that they simply don’t know what they don’t know.
My job, and that of my colleagues, both at Spectra and within the tape industry overall, is to educate as many people as possible about how to reduce the cost, complexity and fear of managing exponentially growing data. Spectra is leading the charge to create an awareness of how valuable tape can now be in the data center. Tape is no longer used just for backup. It was great to see so many of our HPC customers at SC11, most of whom don’t even use the terminology of “backup” any longer. As tape continued to mature over the last 10 years by getting 700% more reliable, faster and more dense, many of our HPC customers started leveraging the benefits of tape in what we call an “Active Archive”. In other words, they are using tape as disk. An active archive is a combination of open system applications, varying types of disk, and tape hardware that intelligently monitors and migrates data across multiple storage devices while maintaining fast user accessibility. Traditionally, in the backup world, one could only access tapes and the data on them through a proprietary backup application such as NetBackup, Legato, Commvault, etc. I’m not advocating that corporations discontinue backups all together because one should always have a “second” copy of data in the event of a disaster. However, the premise of an active archive is that all data can be online all the time.
Obviously, when someone has hundreds of terabytes or even petabytes, it is cost prohibitive to try and keep all data online all the time in the traditional way of keeping it all on primary or secondary disk. With an active archive file system, the data can be dynamically distributed across multiple storage platforms including disk and tape. Policies can determine where data is at any given time and it is transparent to the end user where that might be. They simply have a drive letter and directory with all their files as normal. Nothing proprietary about access to their data—anytime they need it. By extending a file system across high performing disk, capacity disk and now tape, the need for IT intervention to retrieve an archived file is minimized, if not eliminated. This data management approach is being used by many of our HPC customers and they are benefiting tremendously by having a searchable, compliant format to store data for the total lifecycle of a file based on policies, industry regulations and laws.
I could go on about the benefits of active archive or the inherent values that are characteristic of the tape technologies of today, but I would rather provide some links to more information on both so you can continue your own research and put aside any tendencies you might have to subscribe to the “ignorance is bliss” philosophy! Tape is here to stay and is poised to solve your storage headaches today and in the future by offering greater efficiency, better reliability and maximum performance. So wake up! Data Armageddon: tape’s got this one.